Reasearch On Structure Design And Motion Planning Of The Modular Personal Robot

Modular personal robot is an autonomous mobile robot system composed of numbers of modules. Modular personal robot capable of performing basic tasks such as autonomous navigation and localization, grasping of objects, perception of the unknown environment and so on. Compared to the regular robot, they are more convenient to maintenance, repair and update. If one module failure, we can replace it at once without change of other modules, so we can reduce the expense of maintenance sharply. We can also reduce the production cost of modules by mass production. To some complicated tasks which one robot can't perform such as transporting of heavy object and passing step or gap, robots can perform it by connecting and cooperating with each other.In this paper, a novel modular personal robot is designed firstly. The robot is composed of four modules which are driven module, sensor module, connecting module and control module. The structure of the four modules are designed and computed by using PROE and AUTOCAD. The function and composing of modules, the localization method and control structure are also introduced detailed. A new mechanical connecting module is designed to implement the stability and practicability of connecting.Then, the kinematics and dynamics math models of modular personal robot are established, which lays a basis for researching 2D and 3D path-planning algorithm and motion control. The capability of obstacle negotiation is analyzed by establishing single and multi-robots math models. Comparing all kinds of unknown environment path-planning algorithm, we put forward a new fuzzy logical control system based on perception—action. The fuzzy logical controller is established base on this method. Mobotsim 2D simulating proved the validity and reliability of this method.At present, most research of mobile robot path-planning are based on 2D map, rare in 3D environment. In order to validate the control effect of real robot in this method, we choose Webots to complete the complicated unknown environment 3D simulating, the simulating result proved the practicability and portability of this method. Based on multi-robots obstacle negotiation analysis, multi-robots passing step task is simulated using Webots, which proved that some complicated tasks can be completed by multi-robots collaborating.